Extended, multi-agent and multi-objective (MaMoRL) environments based on DeepMind's AI Safety Gridworlds. This is a suite of reinforcement learning environments illustrating various safety properties of intelligent agents. It is made compatible with OpenAI's Gym/Gymnasium and Farama Foundation PettingZoo.

Connect-4 AI inspired by the AlphaZero paper that uses Monte-Carlo Tree Search, and a neural policy and value estimator neural network trained with samples generated from self play between previous iterations of the model.

Console Tic-tac-toe game for playing against the machine or a human.

A nice game of Tetris vs a basic AI

This project includes an analysis of Wordle games published on Twitter, the creation of models that imitate players and metrics used to evaluate them.

This repository consists of a GUI based connect-4 game which will have a AI (implemented using minimax algorithm) playing against a human player.

A chess game and artificial intelligence